the behavior of electrodes ofplatinum and platinum alloys in

BY R. G. VAN NAME AND F. FENWICK'. RECEIVED JUNE 23, 1924. PUBLISHED JANUARY. 8, 1926. It has already been stated in Part I of this article2 that the ...
0 downloads 0 Views 536KB Size
Jan., 1925

POLARIZED ELECTROD& IN ANALYSIS

I9

4. A freshly ignited platinum electrode, immersed in a solution of silver nitrate, usually acquires the electromotive behavior of silver. This is probably due to an invisible coating of that metal. NEWHAVEN, CONNECTICUT [CONTRIBUTION FROM THE

DEPARTMENT OF CHEMISTRY,

YALE UNIVERSITY ]

THE BEHAVIOR OF ELECTRODES OFPLATINUM AND PLATINUM ALLOYS IN ELECTROMETRIC ANALYSIS. 11. POLARIZED ELECTRODES BY R. G. VANNAME AND F. FENWICK' RECEIVED JUNE 23, 1924

PUBLISHED JANUARY 8, 1926

It has already been stated in Part I of this article2 that the electrode system found most useful for determining the end-point in actual titrations consisted of two identical, pure platinum electrodes between which a very low polarizing current was allowed to flow continuously. The latter was obtained by applying a constant potential of from 0.20 to 1.00 volt and interposing a fixed resistance of 20,000 ohms in series with the two electrodes. Polarized electrodes have the great practical advantage of giving more definite and reproducible potentials than those given by the unpolarized electrodes described in the first article. Such a polarized electrode system in contact with a solution of ferrous sulfate exhibits a difference in potential which decreases rapidly on the addition of a small amount of potassium permanganate. Near the completion of the oxidation the difference increases, rises very sharply as the last traces of ferrous ion are oxidized and decreases again as the excess of the oxidizing agent increases. The titration curve has, therefore, the same characteristics as the curve obtained for the same reaction with unpolarized electrodes under the conditions discussed in Part I. That the curve representing the potential difference between the polarized electrodes should resemble in a general way the similar curve for the unpolarized electrodes is only what would be expected, since in both cases the main determining factor is the change in the degree of reversibility of the electrode reactions as the titration progresses. The closeness of the resemblance, however, varies considerably with the nature of the reaction concerned. In the titration of ferrous iron with permanganate, for example, the solution, except a t the very start and near the end-point, contains both ferric and ferrous ion in appreciable concentrations. Since the reversibility of the ferric-ferrous electrode is relatively high, both anodic and cathodic polarization are practically destroyed and the electrodes show National Research Fellow in Chemistry,

* THISJOURNAL, 47, 9 (1925).

20

R. G. VAN NAME AND P. PENWICK

Vol. 47

little or no potential difference. As the end-point is approached ferrous ion disappears and permanganate ion appears and the potentials of both electrodes rise sharply. At this higher potential the depolarizing effect of ferric ion is too slovr and weak to he felt, so that at the end-point, and

just beyon neither electrode is effectively depolarized. With increasing concentration of Permanganate ion, however, the cathodic depolarization becomes fairly effective and as a result of the higher potential the manganous ion acquires the power to depolarize the anode. This accounts for the fact that the potentials tend to draw together again beyond the end-point.

Jan., 1925

POLARIZED GLECTRODES I N ANALYSIS

21

Fig. 1 gives the single potential curves for this reaction in the vicinity of the end-point only, and also the curves showing the difference in potential between the electrodes, (€&-ac). On these and the following "difference curves" the end-point, as judged solely from the shape of the curve, is designated by E. This usage is slightly different from that of Part I.2a In the graphs of the single curves the behavior of the anode is indicated by full lines, the cathode by broken. The sharp fall in potential of the anode beyond the pd-point, observed in this and in other cases in which the polarizing 2 potential was 0.5 volt, or over,3 is probably due to the deposition of a film of manganese dioxide on the anode after all the ferrous ion has been oxidized. A distinct darkening of the electrode was noticeable in these cases and it was proved by special tests that an electrode artificially coated with manganese dioxide gave a lower potential than did bright platinum. The titration of iron with dichromate, Fig. 2, is another example of the same type. The fact that the dichromate electrode is less reversible than the perman+-lCC.-3 __ ganate electrode was shown plainly by pig, 2,-Titration of pe++with the much smaller tendency for the curves K2Cr207. to draw together beyond the end-point. Solution, 0.033M FeS04, 3.8M A reduction reaction in which the reversi- HC1. Titrating solution, 0.017M KzApplied polarization: Graph bility is high before the end-point, and CrL07. 1,O.ov.; Graph 2, o.5 v; is much lower beyond it, is the titration of the differencecurve for Graph 2. ferric iron with trivalent titanium, shown in Fig. 3. In all three of these cases the end-point is marked by a sudden rise in the potential difference between the electrodes. An example of the reverse type, where the electrode reactions are nonreversible before the end-point but reversible beyond it, is given by the titration of thiosulfate with iodine, Fig. 4. I n this case the polarized electrodes show a difference in potential up to the end-point where the difference suddenly falls to zero. The precipitation of zinc ion by potassium ferrocyanide, Fig. 5, is a reaction of approximately this same type because of the presence of a small amount of ferricyanide in the titrating solution. The potential difference decreases sharply at the end-point

c

c

2e

a

THISJOURNAL, 47, 14 (1925).

Fig. 6.

22

R. G. VAN NAME AND F. FENWICK

VOl. 47

but does not disappear entirely. It could, no doubt, be made to disappear by using a mixture of ferro- and ferricyanide as the titrating solution, as has been suggested.* Cases like these just considered, in which the reaction has a high reversibility on one side of the end-point and limited reversibility on the other,

20C

100

0

r/i $

-100

;

.r(

=:

54

-200

-300

5:

E b / - - t

+-lCC.-+ Fig. 3.-Titration of Fe+++with Ti+++. Solution, 0.005M Fez(SO&, 1.2M HnSOh. Titrating solution, 0.03M Ti$(SOc)r. Applied polarization: Graph 1, 0.0 v.; Graph 2, 0.2 v.; Graph 3, 0.5 v; Graph 4 is the difference curve for Graph 2; and Graph 5, that for Graph 3.

make up a considerable fraction of the oxidation-reduction reactions that are use€ul in volumetric analysis and it is evident that with this type of reaction in general polarized electrodes may be expected to give a definite upward or downward break a t the end-point. When the reversibility is limited on both sides of the end-point, the de4

Kolthoff, Rec. brav. chh%, 41, 425 (1922).

Jan., 1925

POLARIZED ELECTRODES I N ANALYSIS

23

termination of the end-point by the use of polarized electrodes may or may not be practicable. Such reactions are frequently incomplete or slow, and therefore available for volumetric analysis only in an indirect method, if at all. The titration of dichromate with sulfite is an example. With polarized electrodes no end-point break is obtained when the sulfite is added slowly, although an apparent end-point, variable in position with the

.+lCC.--, Fig. 4.-Titration of Na&O* with I,. Solution, 0.01M Na&Oa. Titrating solution, 0.05M Iz. Applied polarization: Graph l , O . O v.; Graph 2,0.2 v.; Graph 3, 0.5 v.; Graph 4 is the difference curve for Graph 2; and Graph 5, that for Graph 3.

speed of the titration, may be observed when the addition of the sulfite is sufficiently rapid. In this group, however, there are a number of reactions with which polarized electrodes give sharp end-points. Two of these are shown in Figs. 6 and 7. The oxidation of hydrogen peroxide with permanganate, Fig. 6, is somewhat slow near the end-point, but when the titration is not carried out too rapidly the end-point is definite and reproducible. Fig. 7 shows the titration of tetravalent cerium with trivalent titanium. The relatively poor depolarization of the anode by an excess of the titanous sulfate is

24

R. G. VAN NAME AND P. FENPNCK

VOl. 47

rather surprising. Notwithstanding its strength as a reducing agent it does not undergo anodic oxidation as readily as might be expected. The possible case of a reaction in which the electrode potentials have a high degree of reversibility on both sides of the end-point can be very

+-lCC.-+ Fig. 5.-Titration of Zn++ with Kfie(CN)B. Solution, 0.015M ZnS04, 1.OM NH4C1, 0.6MHC1. Titrating solution, 0.05M KdFe(CN)G. Applied polarization: Graph l: Curve x, 0.0 v.; Curves AI and 1, 0.2 v.; Curves A2 and CZ,0.5 v.; Curves A3 and Ca, 1.0 v. Graph 2 : Curve A1 is the difference; Curve for Ai, C1; Curve A,, for Az, C2; and Curve A,,for As, Ca.

briefly dismissed, for it is doubtful whether a case of this kind exists in which the equilibrium lies near enough to one side to meet the requirements of an analytical method. Moreover, it is evident that in such a case the use of polarized electrodes would be impracticable. Depolarization would be nearly complete throughout the reaction and there would be no appreciable break a t the end-point. A comparison of the form, as shown in the graphs, of the single-poten-

Jan., 1925

25

POLARIZED ELECTRODES IN ANALYSIS

tial curves with and without polarization, shows that in some cases the polarization has increased the magnitude of the end-point break. This

+-lCc.-+ Fig. B.-Titration of H ~ Owith Z KMnO4. Solution, 0.012M HzO~,0.9M H2SOn. Titrating solution, 0.01M KMn04. Applied polarization: Graph 1, 0.0 v.; Graph 2, 0.2 v.; Graph 3, 0.5 v.; Graph 4 is the difference curve for Graph 2; and Graph 5, that for Graph 3.

indicates that polarization (with respect to an auxiliary platinum electrode) will in some cases improve the end-point break in titrations of the ordinary TABLE I Reaction

EFFECT OF POLARIZATION ON THE SINGLE ELECTRODE BREAK Anode Cathode

Fe++-Mn04Fe ++-C!rz07-Fe++ +-Ti +++ SaOa---Ia Zn + +-Fe( CN)a-’-HaOz-MnOaCe+++ +-Ti+ + +

Increased Increased Decreased (more than cathode increased) Much decreased Considerably increased Somewhat decreased Somewhat decreased (negligible, cornpared with total break)

Decreased Decreased Increased Much increased Somewhat decreased Little changed Little changed

type, that is, those based on measurement of the e.m.f. between an inert electrode and a constant normal electrode. This expedient has been found

26

R. G. VAN NAME AND F. PENWICK

VOl. 47

to be of great practical value, both because of the sharper break and because polarized electrodes in general give potentials that are more definite and less subject to annoying variations dependent upon the pretreatment of the electrode. For this reason it is often useful even in cases where the magnitude of the break is not appreciably increased. Table I

C-lCc.+ Fig. 7.-Titration of Ce++++with Ti+++. Solution, 0.006M Ce(SO&, 1.OM HzS04. Titrating solution, 0.03M TidS04)3. Applied polarization: Graph 1, 0.0 v.; Graph 2, 0.2 v.; Graph 3, 0.5 v ; Graph 4 is the difference curve for Graph 2; and Graph 5 , that for Graph 3.

gives a summary of the results of polarization on the single break for the various reactions studied. These results show that polarization has a greater effect upon the single electrode potentials, the more nearly the reaction in question approaches to complete reversibility on one side of the end-point and to complete irreversibility on the other. It is also evident that to increase the magnitude

Jan., 1925

POLARIZED ELECTRODES I N ANALYSIS

I'

L+

n

3

0 I I

Y

: : ? : : : . . $' ' ,

27

28

VOl. 47

R. G. VAN NAME AND .'E FENWICK

of the break the electrode should be polarized in the direction in which depolarization is least effective on the irreversible side of the end-point. Up to a certain point the break can be increased by increasing the polarizing potential, but too high a potential has a tendency to make the bend at the beginning of the break less abrupt and the break less vertical, so this expedient must be used with caution. Displacement of Break by Polarization In consequence of the effect just mentioned, the break with a polarized electrode does not always exactly coincide with the end-point as given by an unpolarized electrode. Thus in an oxidation reaction anodic polarization would tend to make the apparent end-point too early and cathodic polarization to make it too late. The magnitude of this effect would be less, the greater the reversibility in the end-point region. Various instances of such displacement are visible in the graphs and these are summarized in Table 11. The conclusion to be drawn is that the displacement, from a practical standpoint, is not serious and that the error involved in determining the end-point from the titration curve, E,-& is well below the equivalent of 0.05 cc. of 0.1 N titrating solution. Effect of Rate of Stirring The efficiency of depolarization depends, among other factors, upon the concentration of the depolarizer at the surface of the electrode and this in turn upon the efficiency of the stirring. As a rule, therefore, the rate of stirring will have some effect upon the electrode potential. This point was tested with a number of the reactions mentioned above. The potential changes were small in all cases, often unobservable and always negligible as regards their effect upon the end-point break. The most marked results were obtained with a solution approximately 0.001 M with respect to iodine and potassium iodide and 0.9 M with respect t o sulfuric acid. They are given in Table 111. TABLE I11

EFFBCT OF STIRRING Applied potential Volts

1.0 0.5

Ea

Mv.

313 309 303 300

Ea

Mv.

R.p.m.

252 280 252 278

200 600

200 600

Since the depolarization is improved by increased stirring, the two electrode potentials should be affected in opposite directions, as is actually observed.

Jan., 1925

SOLID SOLUTIONS OF HALIDES

29

Summary 1. The principal factor determining the sharpness of the electrometric end-point with polarized electrodes of pure platinum is the difference in the degree of reversibility of the electrode reaction before and after the end-point. Other factors such as concentration and rate of stirring are of minor importance. The end-point phenomena are more distinct, the more nearly the electrode reactions conform to the requirement of complete reversibility on one side of the end-point and complete irreversibility on the other, but are sharp enough for practical use in analysis in some cases which fall far short of meeting this requirement. 2. Although the polarization sometimes produces a slight displacement of the end-point break of a single electrode, the errors so introduced are generally too small to be of significance in a volumetric analysis. 3. The end-point break in titrations of the ordinary type, based on the single potential of a platinum electrode, may often be improved and the behavior of the electrode made much more dependable and satisfactory by polarizing it with respect to an auxiliary electrode. NEWHAVEN,CONNECTICUT [CONTRIBUTION FROM THE PHYSICAL AND CHEMICAI, LABORATORIES OF THE OHIO STATE UNIVERBITY]

SOLID SOLUTIONS OF THE ALKALI AND AMMONIUM HALIDES BY R. J. HAVIGHURST, E. MACK,JR. AND I?. C.BLAKE RECEIVED JULY 16, 1924

PUBLISHED JANUARY 8, 1925

The purpose of this paper is to establish the limits of miscibility of the alkali and ammonium halides a t room temperature and to determine the natures and structures of their solid solutions. The photographic powderedcrystal method of X-ray analysis was employed. The apparatus and method of procedure for precision work have been described.l The constitution of metallic alloys has been frequently investigated by the X-ray method, but Vegard2 seems to have been the only one to have used this procedure in the investigation of solid solutions of salts. The problem of the miscibility of the alkali halides has often been studied by thermal and optical methods. These methods sometimes lead to ambiguous results, and are not always sufficient to permit definitely deciding whether a pair of salts is miscible a t room temperature. X-ray methods, on the other hand, are capable of giving a definite answer to this question, provided only that the system be in equilibrium. The Nature of Solid Solution.-Solid solution may take place in two different ways. The mechanism of the simpler of these two ways is a more Havighurst,

a Vegard,

Mack and Blake, THISJOURNAL, 46,2368 (1924).

Z.Physik, 5, 17 (1921).